http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
탈중앙거래소의 비영구적 손실 제거를 위한 딥러닝 기반 Dynamic AMM
김수민(Sumin Kim),최해웅(Haeung Choi),강주성(Jusung Kang),윤민호(Minho Yoon),이흥노(Heung-No Lee) 대한전자공학회 2023 대한전자공학회 학술대회 Vol.2023 No.6
Decentralized exchanges (DEXs), based on blockchain technology, enable the exchange of various cryptocurrencies. A common problem in DEXs is Impermanent Loss (IL), which forces users of DEX to experience excessively large slippage or transaction delays. To tackle this problem, a Dynamic Curve Automated Market Maker (AMM) structure was proposed [2]. However, this method requires an infeasible amount of transaction costs for real-time market price oracle. In this study, we propose a deep learning-based Dynamic AMM to mitigate IL. The proposed method utilizes predicted prices instead of real-time prices to alleviate the IL problem in DEXs and to reduce transaction costs to a practical level. Through simulation, we have validated that this approach can effectively decrease IL and make transaction costs more manageable in the context of DEXs.